Estimation of the number of synapses in the hippocampus ...Feb 18, 2020 · 1 1 Estimation of the number of synapses in the hippocampus and 2 brain-wide by volume electron microscopy

1

Estimation of the number of synapses in the hippocampus and 1

brain-wide by volume electron microscopy and genetic labeling 2

3

Andrea Santuy1, Laura Tomás-Roca2, José-Rodrigo Rodríguez1,3,4, Juncal González-Soriano5, 4 Fei Zhu2,6, Zhen Qiu2, Seth GN Grant2, Javier DeFelipe1,3,4, Angel Merchan-Perez1,4,7 5

6

1 Laboratorio Cajal de Circuitos Corticales, Centro de Tecnología Biomédica, Universidad 7 Politécnica de Madrid, Pozuelo de Alarcón, 28223 Madrid, Spain. 8 2 Genes to Cognition Program, Centre for Clinical Brain Sciences, University of Edinburgh, 9 Edinburgh EH16 4SB, UK 10 3 Instituto Cajal, Consejo Superior de Investigaciones Científicas (CSIC), Avda. Doctor Arce, 37, 11 28002 Madrid, Spain. 12 4 Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas 13 (CIBERNED) ISCIII, Madrid, Spain. 14 5 Departamento de Anatomía y Embriología, Universidad Complutense de Madrid, 28040 15 Madrid, Spain. 16 6 UCL Institute of Neurology, Queen Square, WC1N 3BG London, UK 17 7 Departamento de Arquitectura y Tecnología de Sistemas Informáticos, Universidad 18 Politécnica de Madrid, Pozuelo de Alarcón, 28223 Madrid, Spain. Corresponding author 19 20 JDF, SG and AMP designed research; AS, LTR, JRR, JGS, FZ and ZQ performed research; AS, LTR, 21 and AMP analyzed data; AS, LTR, SG, JDF and AMP wrote the paper. 22

23

Abbreviated Title: Brain-wide estimation of synapses 24

25

Correspondence should be addressed to: Angel Merchan-Perez, Laboratorio Cajal de Circuitos 26 Corticales, Centro de Tecnología Biomédica, Universidad Politécnica de Madrid, Pozuelo de 27 Alarcón, 28223, Madrid, Spain: [email protected] 28

29

30

31

32

33

.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is

The copyright holder for this preprintthis version posted April 28, 2020. ; https://doi.org/10.1101/2020.02.18.953802doi: bioRxiv preprint

https://doi.org/10.1101/2020.02.18.953802

http://creativecommons.org/licenses/by-nc-nd/4.0/

2

Abstract 34

Membrane-associated guanylate kinases (MAGUKs) are a family of scaffolding proteins that 35 are expressed in excitatory glutamatergic synapses. We used genetic labeling of two of these 36 proteins (PSD95 and SAP102), and Spinning Disc confocal Microscopy (SDM), to estimate the 37 number of fluorescent puncta in the CA1 area of the hippocampus. We also used FIB-SEM, a 38 three-dimensional electron microscopy technique, to calculate the actual numbers of synapses 39 in the same area. We then estimated the ratio between the three-dimensional densities 40 obtained with FIB-SEM (synapses/µm3) and the bi-dimensional densities obtained with SDM 41 (puncta/100 µm2). Given that it is impractical to use FIB-SEM brain-wide, we used previously 42 available SDM data from other brain regions and we applied this ratio as a conversion factor to 43 estimate the density of synapses. We found the highest densities of synapses in the isocortex, 44 olfactory areas, hippocampal formation and cortical subplate. Low densities were found in the 45 pallidum, hypothalamus, brainstem and cerebellum. Finally, the striatum and thalamus 46 showed a wide range of synapse densities. 47

48

Keywords 49

FIB-SEM, PSD95, SAP102, Hippocampus, CA1 50

51

Significance statement 52

Determining the number of synapses that are present in different brain regions is crucial to 53 understand brain connectivity as a whole. Here we have developed a technique that allows us 54 to calculate the number of synapses in the hippocampus, and to estimate the minimum 55 number of excitatory synapses that are present in more than one hundred regions brain-wide. 56 We found that, in general, rostral parts of the brain —including the neocortex— have the 57 highest numbers of synapses, while caudal regions —including the cerebellum and 58 brainstem— have much lower numbers. Other regions, such as the thalamus, are highly 59 heterogeneous. These quantitative data will help us to better understand the structure of 60 brain microcircuits and to build realistic whole-brain models. 61

62

63



https://doi.org/10.1101/2020.02.18.953802


3

Introduction 64

Membrane-associated guanylate kinases (MAGUKs) are a family of scaffolding proteins that 65 participate in the regulation of cell polarity, cell adhesion and synaptic signal transduction 66 (Migaud et al., 1998; Ye et al., 2018; Zhu et al., 2016). PSD95 and SAP102 belong to the 67 MAGUK family and are expressed in the postsynaptic density (PSD) of excitatory glutamatergic 68 synapses (Aoki et al., 2001; Chen et al., 2018, 2008; DeGiorgis et al., 2006; Farley et al., 2015; 69 Husi et al., 2000; Petersen et al., 2003; Valtschanoff et al., 1999; Yamasaki et al., 2016), where 70 they contribute to the recruitment and retention of glutamate receptors (Hafner et al., 2015; 71 Jeyifous et al., 2016; Levy et al., 2015). Genetic labeling of the endogenous PSD95 and SAP102 72 postsynaptic proteins and imaging using Spinning Disk confocal Microsocpy (SDM) have been 73 proven to be useful for the characterization of synapse diversity in all brain regions of the 74 mouse. High-speed SDM allows the imaging of entire brain sections and visualization of 75 millions of synapses at a time, providing bi-dimensional densities of fluorescent puncta per 76 surface area (puncta/100 µm2) (Zhu et al., 2018). 77

Previous attempts have been made to calculate the density of synapses in the brain using 78 electron microscopy (EM). This technique allows the identification of individual synapses, 79 although it is restricted to much smaller fields of view. Furthermore, most of these EM studies 80 apply stereological techniques to a limited number of EM sections. Although stereology is a 81 proven valuable method for object counting, the total number of synapses is an estimation 82 which is subject to several technical limitations [see (DeFelipe et al., 1999) for a review]. In the 83 present study, we use Focused Ion Beam milling-Scanning Electron Microscopy (FIB-SEM). With 84 this technique, sectioning and imaging are fully automated, allowing the acquisition of multiple 85 serial micrographs. Later, the micrographs can be stacked with the help of software tools, such 86 that they represent a three-dimensional sample of tissue (Merchán-Pérez et al., 2009). In this 87 way, all individual synapses can be identified and counted within a known volume of brain 88 tissue, and thus the true density of synapses per unit volume can be obtained directly (not 89 through estimations using stereological methods). 90

The aim of our study was twofold. First, we wanted to obtain detailed data about the density 91 and size of synapses in the hippocampus. To this end, we used SDM to measure the densities 92 of PSD95 and SAP102 puncta in stratum oriens (SO), stratum radiatum (SR) and stratum 93 lacunosum-moleculare (SLM) of CA1. We also used FIB-SEM to measure the actual density of 94 synapses in three-dimensional samples of the same strata (Figure 1). We then calculated the 95 quantitative relationship between the densities and sizes of fluorescent puncta and synapses 96 obtained by the two methods. Second, given that volume electron microscopy cannot be 97 applied brain-wide, we wanted to obtain an estimate of the number of synapses in other 98 regions of the brain where measurements of PSD95 and SAP102 puncta were available (Zhu et 99 al., 2018). We based this estimate on the quantitative relationship or conversion factor 100 between SDM and FIB-SEM data previously obtained in the hippocampus. Even though this 101 approach has several limitations and underestimates the actual numbers of synapses, it 102 provides valuable information on the minimum number of excitatory synapses that are present 103 in more than a hundred brain regions. 104

105

106

107



https://doi.org/10.1101/2020.02.18.953802


4

108

109

110

Figure 1. General methodology. Knockin mice expressing fluorescent PSD95 and SAP102 were 111 imaged with Spinning Disc Confocal Microscopy (SDM) and with FIB-SEM. SDM allows the 112 acquisition of large field, 2D fluorescent images, while FIB-SEM is an electron microscopy 113 technique with a resolution in the scale of nanometres that generates 3D stacks of images, but 114 with a smaller field of view. We have used a combination of both techniques to estimate the 115 actual densities of synapses per unit volume brainwide. 116

117

118



https://doi.org/10.1101/2020.02.18.953802


5

Materials and Methods 119

Animals 120

For this study, we used adult male mice (postnatal day 56) expressing fluorescently labeled 121 PSD95 and SAP102 postsynaptic proteins (PSD95eGFP/eGFP; SAP102mKO2/Y) (Zhu et al., 2018). All 122 animals were handled in accordance with the guidelines for animal research set out in the 123 European Community Directive 2010/63/EU, and all procedures were approved by the local 124 ethics committee. 125

Tissue preparation for spinning disc microscopy 126

Sixteen mice were anesthetized by an intraperitoneal injection of 0.1 mL of 20% w/v sodium 127 pentobarbital (Euthatal, Merial Animal Health Ltd. or Pentoject, Animalcare Ltd.). After 128 complete anesthesia, 10 mL of phosphate buffered saline (PBS; Oxoid) were perfused 129 transcardially, followed by 10 mL of 4% v/v paraformaldehyde (PFA; Alfa Aesar). Whole brains 130 were dissected out and post-fixed for 3–4 h at 4° C in 4% PFA, and then cryoprotected for 3 131 days at 4 °C in 30% sucrose solution (w/v in 1× PBS; VWR Chemicals). Brains were then 132 embedded into optimal cutting temperature (OCT) medium within a cryomould and frozen by 133 placing the mould in isopentane cooled down with liquid nitrogen. Brains were then sectioned, 134 with a thickness of 18 μm, using an NX70 Thermo Fisher cryostat, and cryosections were 135 mounted on Superfrost Plus glass slides (Thermo scientific) and stored at −80 °C. 136

Histology and immunohistochemistry 137

Sections were washed for 5 min in PBS, incubated for 15 min in 1 µg/mL DAPI (Sigma), washed 138 and mounted using home-made MOWIOL (Calbiochem) containing 2.5% anti-fading agent 139 DABCO (Sigma-Aldrich), covered with a coverslip (thickness #1.5, VWR international) and 140 imaged the following day. 141

Spinning Disk Confocal Microscopy 142

For synaptome mapping, we used Spinning Disk confocal Microscopy (SDM) platforms. The 143 Andor Revolution XDi was used with an Olympus UPlanSAPO 100X oil immersion lens (NA 1.4), 144 a CSU-X1 spinning-disk (Yokogawa) and an Andor iXon Ultra monochrome back-illuminated 145 EMCCD camera. Images acquired with this system have a pixel dimension of 84 × 84 nm and a 146 depth of 16 bits. A single mosaic grid was used to cover each entire brain section with an 147 adaptive Z focus set-up by the user to follow the unevenness of the tissue using Andor iQ2 148 software. In both systems, eGFP was excited using a 488 nm laser and mKO2 with a 561 nm 149 laser. The CV1000 system is equipped with the following filters: BP 525/50 nm for eGFP and BP 150 617/73 nm for mKO2, whereas the Andor Revolution XDi is equipped with a Quad filter (BP 151 440/40, BP 521/21, BP 607/34 and BP 700/45). For both systems, mosaic imaging was set up 152 with no overlap between adjacent tiles. 153

Detection and measurement of fluorescent Synaptic Puncta 154

Punctum detection was performed using Ensemble Detection, an in-house collection of image 155 detection algorithms. We have developed a new punctum/particle detection method based on 156 a multi-resolution image feature detector and supervised machine learning technique (Zhu et 157 al., 2018). In this method, we carry out a multi-resolution and multi-orientation version of 2nd-158 order nonlocal derivative (NLD) (Qiu et al., 2012), and use it to calculate intensity differences, 159 referred to as ‘image features’, for each of the individual puncta at different spatial resolutions 160



https://doi.org/10.1101/2020.02.18.953802


6

and orientations. An initial intensity threshold is set to a very low value to only filter out 161 extremely dim puncta and to avoid missing true synaptic puncta. The remaining candidate 162 puncta were finally classified as either true puncta or background noise using the 163 corresponding feature vectors and the classifier. The classifier was pre-trained with the 164 training image set and machine learning algorithms (Qiu et al., 2012). 165

After detection and localization of all puncta, we segmented them based on their individual 166 intensity values: for each punctum, a threshold was set as 10% of the maximum pixel intensity 167 within the punctum, so that punctum size and shape measurement were independent of 168 punctum intensity (Zhu et al., 2018). With the puncta segmented and binarized, six punctum 169 parameters were then calculated: mean punctum pixel intensity, punctum size, skewness, 170 kurtosis, circularity, and aspect ratio. 171

Tissue Preparation for electron microscopy 172

Four male PSD95eGFP/eGFP; SAP102mKO2/Y mice were used for electron microscopy. Animals were 173 administered a lethal intraperitoneal injection of sodium pentobarbital (40 mg/kg) and were 174 intracardially perfused with 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M 175 phosphate buffer (PB). The brain was then extracted from the skull and processed for EM as 176 previously described (Merchán-Pérez et al., 2009). Briefly, the brains were post-fixed at 4°C 177 overnight in the same solution used for perfusion. They were then washed in PB and 178 vibratome sections (150 μm thick) were obtained. Sections containing the rostral hippocampus 179 were selected with the help of an atlas (Paxinos and Franklin, 2004). Selected sections were 180 osmicated for 1 hour at room temperature in PB with 1% OsO4, 7% glucose and 0.02 M CaCl2. 181 After washing in PB, the sections were stained for 30 min with 1% uranyl acetate in 50% 182 ethanol at 37°C, and they were then dehydrated and flat embedded in Araldite (DeFelipe and 183 Fairén, 1993). Embedded sections were glued onto blank Araldite stubs and trimmed. To select 184 the exact location of the samples, we first obtained plastic semithin sections (1−2 μm thick) 185 from the block surface and stained them with toluidine blue to identify cortical layers. These 186 sections were then photographed with a light microscope. The last of these light microscope 187 images (corresponding to the section immediately adjacent to the block face) was then 188 collated with scanning electron microscope (SEM) photographs of the surface of the block. In 189 this way, it was possible to accurately identify the regions of the neuropil to be studied. 190

Three-Dimensional Electron Microscopy 191

Three-dimensional brain tissue samples of the CA1 of the hippocampus were obtained using 192 combined focused ion beam milling and scanning electron microscopy (FIB-SEM) (Figure 2). 193 The focus of our study was the neuropil, which is composed of axons, dendrites and glial 194 processes. We used a CrossBeam 540 electron microscope (Carl Zeiss NTS GmbH, Oberkochen, 195 Germany). This instrument combines a high-resolution field emission SEM column with a 196 focused gallium ion beam, which can mill the sample surface, removing thin layers of material 197 on a nanometer scale. After removing each slice (20 nm thick), the milling process was paused, 198 and the freshly exposed surface was imaged with a 1.8-kV acceleration potential using the in-199 column energy selective backscattered (EsB) electron detector. The milling and imaging 200 processes were sequentially repeated, and long series of images were acquired through a fully 201 automated procedure, thus obtaining a stack of images that represented a three-dimensional 202 sample of the tissue (Merchán-Pérez et al., 2009). Twelve samples (stacks of images) of the 203 neuropil of three strata of CA1 were obtained, avoiding the neuronal and glial somata as well 204 as the blood vessels (Figure 3). These stacks included four samples of stratum lacunosum 205



https://doi.org/10.1101/2020.02.18.953802


7

moleculare (SLM), four of stratum radiatum (SR) and four of stratum oriens (SO) (see 206 Supplementary Table 1). In these stacks, we obtained the densities of glutamatergic 207 (asymmetric) and GABAergic (symmetric) synaptic junctions. To do this, we counted the 208 number of synaptic junctions within an unbiased three-dimensional counting frame of known 209 volume (Howard and Reed, 2005). Image resolution in the xy plane was 5 nm/pixel; resolution 210 in the z-axis (section thickness) was 20 nm and image sizes were 2048 x 1536 pixels. The 211 number of sections per stack ranged from 201 to 377 (mean 276.33; total 3316 sections). 212 Processing for EM causes shrinkage of the tissue for which we have to correct the 213 measurements (Merchán-Pérez et al., 2009). Correction factors were 0.9508 for linear 214 measurements, 0.9040 for area measurements and 0.8595 for volumetric data. The volumes of 215 the stacks, after correction for tissue shrinkage, ranged from 367.81 to 689.86 μm3 (mean 216 505.66 μm3; total 6067.86 μm3). The volumes of the counting frames ranged from 288.62 to 217 585.99 μm3 (mean 408.76 μm3; total 4905.07 μm3) (Supplementary Table 1). 218

Identification and reconstruction of synapses 219

Synaptic junctions within these volumes were visualized and segmented in 3D with Espina 220 software (Morales et al., 2011) (http://cajalbbp.es/espina/). The segmentation algorithm 221 makes use of the fact that presynaptic and postsynaptic densities appear as dark, electron-222 dense structures under the electron microscope. It requires a Gaussian blur filter 223 preprocessing to eliminate noisy pixels and then it uses a gray-level threshold to extract all the 224 voxels that fit the gray levels of the synaptic junction. In this way, the resulting 3D 225 segmentation includes both the active zone (AZ) and postsynaptic density (PSD) (Morales et 226 al., 2013). Synaptic junctions with a prominent or thin PSD were classified as asymmetric or 227 symmetric synaptic junctions, respectively (Colonnier, 1968; Gray, 1959) (Figure 2). Synapses 228 could be unambiguously identified since they can be visualized in consecutive serial sections 229 and, if necessary, they can be digitally resectioned in different planes to ascertain their identity 230 as asymmetric or symmetric synapses (DeFelipe et al., 1999; Merchán-Pérez et al., 2009). 231

Size of synapses 232

As stated above, the synaptic junction is formed by the AZ and the PSD. Since AZ and PSD are 233 in close apposition and have similar surface areas, they can be represented as a single surface 234 — the synaptic apposition surface (SAS). Thus, the SAS is an accurate measurement of the size 235 of the synapse. In previous studies we have developed an efficient computational technique to 236 automatically extract this surface from reconstructed synapses (Morales et al., 2013). 237

Statistical analysis 238

To study whether there were significant differences between synaptic distributions among the 239 different CA1 layers, we performed a multiple mean comparison test. When the data met the 240 criteria of normality and homoscedasticity, an ANOVA was performed. When these criteria 241 were not met, we used the Kruskal-Wallis followed by Dunn’s test for pair-wise comparisons. 242

243

244



http://cajalbbp.es/espina/�

https://doi.org/10.1101/2020.02.18.953802


8

Results 245

We estimated the density and the size of synapses in the hippocampus using two different 246 methods (Figure 1). PSD95-positive and SAP102-positive synapses were identified as 247 fluorescent puncta using SDM, and FIB-SEM was used to visualize and reconstruct synaptic 248 junctions in the same regions. FIB-SEM also provided information that was not obtained from 249 confocal images, such as the relative proportions of excitatory (asymmetric) and inhibitory 250 (symmetric) synapses (Figure 2, Supplementary Table 1). This classification of synapses is 251 based on the appearance of the PSD in EM images (Colonnier, 1968; Gray, 1959). Any synaptic 252 junction with a dense, prominent PSD that was much thicker than the relatively faint 253 presynaptic thickening was classified as “asymmetric” (AS). Any synapse with a less marked 254 PSD, similar to the presynaptic thickening, was classified as “symmetric” (SS) (Merchán-Pérez 255 et al., 2009). It should be stressed that the classification of synaptic junctions into one of these 256 two groups was not based on the examination of single sections, but on the whole series of 257 images in which the PSD was visible (Figure 2). Once all synapses within a given stack of serial 258 sections had been identified and segmented, they appeared as a cloud of 3D objects from 259 which quantitative data were obtained (Figure 3). 260

261

Density of fluorescent puncta and synapses in the hippocampus 262

Densities of fluorescent puncta (number of positive puncta per 100 µm2) were measured in 263 SLM, SR and SO from CA1. Sixteen brain sections from sixteen mice were used (Supplementary 264 Table 2). We obtained the densities of puncta expressing PSD95 (dPSD95), SAP102 (dSAP102) 265 and the colocalization index (c). The colocalization index ranges from 0, when there is no 266 colocalization, to 1, when there is 100% colocalization. From these data we calculated the total 267 density of puncta (dTotal) (Table 1). Note that the total density of puncta is not simply the sum 268 of dPSD95 and dSAP102, since there is a certain density of puncta that colocalize (dColoc). We 269 operated as follows: 270

𝑑𝑇𝑜𝑡𝑎𝑙 = 𝑑𝑃𝑆𝐷95 + 𝑑𝑆𝐴𝑃102 − 𝑑𝐶𝑜𝑙𝑜𝑐

𝑑𝐶𝑜𝑙𝑜𝑐 = 𝑐 × 𝑑𝑇𝑜𝑡𝑎𝑙

therefore, 271

𝑑𝑇𝑜𝑡𝑎𝑙 = 𝑑𝑃𝑆𝐷95 + 𝑑𝑆𝐴𝑃102

1 + 𝑐

We observed the highest density of puncta in SR (mean ± SD; 147.30 ± 27.49 puncta/100 µm2), 272 followed by SO (141.13 ± 17.39 puncta/100 µm2) and SLM (104.46 ± 23.57 puncta/100 µm2) 273 (Table 1). The differences between SLM and the other two layers were statistically significant 274 (KW test, p < 0.005) (Figure 4A). 275

276



https://doi.org/10.1101/2020.02.18.953802


9

277

Figure 2. Identification and segmentation of synaptic junctions in serial sections acquired by FIB-SEM. 278 (A-C) Detail of four electron micrographs selected from a series o images obtained by FIB-SEM. In this 279 example, the stack of images was obtained from the stratum oriens. The numbers in the bottom-right 280 corner correspond to section number. Four asymmetric synapses can be identified by the presence of 281 prominent post-synaptic densities in A, B, C and D (asterisks). One symmetric synapse, with a thin post-282 synaptic density, can be seen in C and D (arrow heads). Note that the classification of synapses as 283 asymmetric or symmetric is not based on single images but on the examination of the full sequence of 284 images. (E-H) The same images after they have been segmented with Espina software 285 (http://cajalbbp.es/espina/). The segmentation process is based on grey-level thresholds, so the 286 resulting 3D objects comprise both the pre- and post-synaptic densities (see methods). Green profiles 287 correspond to asymmetric synapses and red profiles to the symmetric synapse. (I) 3D rendering of the 288 synaptic junctions present in A to H. (J) Synaptic apposition surfaces (SAS, yellow) extracted from the 3D 289 segmentations represented in (I). SAS are automatically extracted from the 3D reconstructions of 290 synaptic junctions (see methods); they are zero-volume surfaces that represent the interface between 291 the pre- and post-synaptic densities. The surface area of the SAS is measured for each individual 292 synaptic junction. (K, L) Same structures represented in I and J, respectively, after they have been 293 rotated through a vertical axis. Original images were acquired with a resolution of 5 nm/pixel, with a 294 distance of 20 nm between two consecutive images. Calibration bar in H is 1 µm. 295



http://cajalbbp.es/espina/�

https://doi.org/10.1101/2020.02.18.953802


10

296

Figure 3. Measuring synaptic densities and sizes in stacks of sections obtained by FIB-SEM. (A-C) 297 Panoramic view of electron micrographs of the stratum lacunosum moleculare (SLM), stratum radiatum 298 (SR) and stratum oriens (SO) imaged by FIB-SEM. (D-F) 3D rendering of synaptic junctions reconstructed 299 from the corresponding stacks of serial sections, acquired from the strata represented in A to C. 300 Asymmetric synaptic junctions have been represented in green and symmetric synaptic junctions in red. 301 (G-I). The synaptic apposition surfaces (SAS, yellow) have been automatically extracted from the three-302 dimensionally reconstructed synaptic junctions. The number of synapses per unit volume and the 303 surface areas of the SAS have been measured in each stack of serial sections (see Tables 1 and 2). 304

305



https://doi.org/10.1101/2020.02.18.953802


11

306

For the volume electron microscopy study (FIB.SEM), we used 12 stacks of serial sections from 307 SLM, SR and SO (Figure 2, Figure 3, Supplementary Table 1). In these samples, we identified 308 and analyzed a total of 10,460 synapses in 4,905 µm3 of tissue. Of these, 95.60% were AS and 309 4.40% were SS. To estimate the density of synapses in each stack of images, we counted the 310 number of synaptic junctions within an unbiased three-dimensional counting frame of known 311 volume (see Methods). The density of AS (mean ± SD) in SLM was 1.59 ± 0.73 synapses/µm3, in 312 SR it was 2.31 ± 0.38 synapses/µm3, and in SO it was 2.49 ± 0.48 synapses/µm3 (Table 1). 313

314

315

Stratum of CA1

Density of PSD95

(puncta/100 µm2)

Density of SAP102

(puncta/100 µm2)

Coloca-lization index

Total density of

puncta/100 µm2

Density of AS

(synapses/µm3)

Conver-sion

factor

Lacunosum-Moleculare

86.9174 ± 19.1856

79.5983 ± 26.1547 0.5973 104.4573 ±

23.5698 1.5958 ± 0.7317 0.0153

Radiatum 122.9793 ± 25.4202

121.0397 ± 22.3355 0.6607 147.3026 ±

27.4920 2.3076 ± 0.3788 0.0157

Oriens 121.1208 ± 19.1566

111.7809 ± 17.4686 0.6520 141.1321 ±

17.3944 2.4887 ± 0.4763 0.0176

All layers averaged

110.3392 ± 26.8614

104.1396 ± 28.2323 0.6367 130.9640 ±

29.6704 2.1307 ± 0.6396 0.0162

316

Table 1. Relationship between the densities of puncta and the densities of synapses. PSD95 and 317 SAP102 puncta were imaged with SDM. The total densities of puncta were calculated from the densities 318 of PSD95 and SAP102 puncta, together with the colocalization index (see text for details). The densities 319 of asymmetric synapses (AS) were obtained from volumes of tissue reconstructed from serial sections 320 using FIB-SEM. For each layer, the conversion factors is the quotient between the density of synapses 321 obtained by FIB-SEM and the total density of puncta calculated from SDM images. Densities are given as 322 average ± SD 323

324

SS were most frequent in SLM (0.13 ± 0.07 synapses/µm3), followed by SR (0.06 ± 0.02 325 synapses/µm3) and SO (0.05 ± 0.02 synapses/µm3). In spite of this trend of an increase in AS 326 density from SLM to SO, and a decrease in SS across these strata, the differences between 327 layers were not statistically significant for either the total density of synapses (AS+SS) or for AS 328 and SS separately (KW test, p ≥ 0.08) (Figure 4B, C). 329



https://doi.org/10.1101/2020.02.18.953802


12

Figure 4. Total densities of PSD95 and SAP102 puncta, asymmetric synapses and symmetric synapses. 330 (A) Density of PSD95- and SAP102-positive puncta (puncta/100 µm2 ± SD) acquired by SDM in the 331 hippocampus (CA1). Asterisks indicate statistically significant differences (KW, p < 0.005). (B) and (C) 332 Density of asymmetric and symmetric synapses, respectively (synapses/µm3 ± SD), estimated from 333 stacks of serial sections acquired by FIB-SEM from the same regions. See also Table 1. 334

335

Size of fluorescent puncta and synapses in the hippocampus 336

We measured the area of PSD95 puncta in SLM, SR and SO. The largest mean area of puncta 337 was found in SLM (0.0832 µm2), followed by SR (0.0809 µm2) and SO (0.0798 µm2). Although 338 the differences were small, they were statistically significant (KW test p < 0.001). 339

To estimate the size of synapses in the FIB-SEM samples, we measured the area of the synaptic 340 apposition surface (SAS). The SAS is a surface that represents the apposition between the 341 presynaptic density and the PSD and reproduces their curvature (Figure 2, Figure 3, see 342 Methods). The mean SAS area in the FIB-SEM samples was 0.0474 µm2 for asymmetric 343 synapses and 0.0541 µm2 for symmetric synapses. 344

For asymmetric synapses, the mean SAS area in SLM was larger than in the other layers 345 (0.0633 µm2; KW-Dunn’s p < 0.001). Despite the mean SAS area being larger in SR than in SO 346 (0.0456 µm2 and 0.0419 µm2, respectively), the difference was not statistically significant (KW-347 Dunn’s p > 0.05). For symmetric synapses, the largest mean SAS areas were found in SO 348 (0.0644 µm2) followed by SLM (0.0519 µm2) and SR (0.0501 µm2) (KW, p = 0.05) (Figure 5, 349 Table 2). 350



https://doi.org/10.1101/2020.02.18.953802


13

351

Figure 5. Size of PSD95 puncta, asymmetric synapses and symmetric synapses. (A) Area of PSD95-352 positive puncta (µm2) acquired by SDM in the hippocampus (CA1) (Mean + SD). Asterisks indicate 353 statistically significant differences. (B) and (C) Mean size of the synaptic apposition surface (SAS) of 354 asymmetric and symmetric synapses (µm2) estimated from stacks of serial sections acquired by FIB-SEM 355 from the same region. See also Table 2. 356

357

When we compared the sizes of asymmetric synapses and symmetric synapses in different 358 layers, we found that symmetric synapses were larger than asymmetric synapses in SO and SR, 359 while in SLM the opposite was the case. The greatest differences were found in SO, where 360 mean SAS areas for symmetric synapses and asymmetric synapses were in a proportion of 361 approximately 6:4 (MW test, p < 0.0001) (Table 2). 362

To further characterize the size distribution of synaptic sizes, we plotted the frequency 363 histograms of the areas of PSD95 puncta and of the SAS (Figure 6). The frequency histograms 364 of the areas of PSD95 puncta showed skewed shapes, with a long tail to the right. The 365 histograms of SAS areas of asymmetric synapses measured from FIB-SEM reconstructions also 366 showed skewed shapes, but they were narrower and lay to the left of PSD95 histograms in all 367 layers (Figure 6). 368

We then performed goodness-of-fit tests to find the theoretical probability density functions 369 that best fitted the empirical distributions of the areas of PSD95 puncta and SAS areas. We 370 found that they fitted to log-normal distributions in all cases, with some variations in the 371 parameters µ and σ (Table 2 and Figure 6). 372

373



https://doi.org/10.1101/2020.02.18.953802


14

374

Table 2. Surface areas of PSD95 puncta acquired by SDM, and surface areas of the synaptic apposition 375 surface (SAS) of asymmetric (AS) and symmetric (SS) synapses reconstructed from FIB-SEM samples. 376 The number of puncta or synapses analyzed (n), as well as the parameters µ and σ of the corresponding 377 best-fit log-normal distributions are also indicated. 378

Figure 6. Frequency histograms of the sizes of PSD95 puncta and asymmetric synapses. (A-C) 379 Comparison of the distribution of the surface areas of PSD95 puncta acquired by SDM (red line) and the 380 synaptic apposition surfaces of asymmetric synapses reconstructed from FIB-SEM samples (blue line) 381 from three layers of the CA1 region of the hippocampus. The histograms corresponding to asymmetric 382 synapses are narrower and lie to the left of the histograms corresponding to PSD95 puncta. (D-F) 383 Frequency histograms (blue bars) of the areas of the synaptic apposition surfaces of asymmetric 384 synapses reconstructed from FIB-SEM samples. The log-normal distributions (black lines) represent the 385 theoretical probability density functions that best fit the experimental data. The parameters µ and σ of 386 the corresponding log-normal distributions have also been indicated. 387

Stratum

Lacunosum-Moleculare Radiatum Oriens All layers

PSD 95 Puncta

Mean Area ± SD (µm2) 0.0832 ± 0.0342 0.0809 ± 0.0352 0.0798 ± 0.0336 0.0813 ± 0.0342

n 147776 189945 125916 463637 µ 11.24 11.20 11.20 11.21 δ 0.44 0.48 0.46 0.46

AS

Mean SAS Area ± SD (µm2) 0.0633 ± 0.0540 0.0456 ± 0.0358 0.0419 ± 0.0271 0.0474 ± 0.0377

n 2258 4538 5082 11878 µ 10.74 10.55 10.50 10.54 δ 0.80 0.63 0.57 0.64

SS

Mean SAS Area ± SD (µm2) 0.0520 ± 0.0338 0.0501 ± 0.0394 0.0644 ± 0.0491 0.0541 ± 0.389

n 247 93 87 427 µ 10.67 10.60 10.97 10.70 δ 0.68 0.65 0.75 0.69



https://doi.org/10.1101/2020.02.18.953802


15

388

Brain-wide estimations of the number of synapses 389

When we compared the densities of PSD95 and SAP102 puncta measured with SDM and the 390 densities of AS measured with FIB-SEM, we found that both methods revealed that SR and SO 391 had similar densities, while SLM had a lower density (Figure 4). We then calculated a 392 conversion factor that would allow us to relate the densities of PSD95 and SAP102 puncta 393 (puncta/100 µm2) to the actual densities of excitatory synapses found by FIB-SEM 394 (synapses/µm3). These conversion factors were calculated as the quotient between the actual 395 density of AS and the total density of PSD95 and SAP102 puncta. Conversion factors obtained 396 for each layer of CA1 were slightly different; they ranged from 0.0152 to 0.0176. The averaged 397 conversion factor calculated with data from the three layers was 0.0162 (Table 1). 398

The next step was to calculate the total number of puncta expressing PSD95 and/or SAP102 399 brain-wide, using previously published data from 113 areas (Zhu et al., 2018). Different brain 400 regions had different combinations of densities of PSD95, SAP102 and total densities of puncta 401 (Figure 7, Supplementary Table 2). The highest total densities of puncta were found in the 402 isocortex, the olfactory areas, the hippocampal formation and the cortical subplate. All these 403 regions were relatively homogeneous except for the hippocampal formation, which showed 404 wider ranges of variability. Also, isocortical areas had a relatively higher proportion of PSD95 405 versus SAP102 than the other regions. The pallidum, the hypothalamus, the brainstem and the 406 cerebellum had low densities of puncta. Finally, the striatum —and especially the thalamus— 407 showed the greatest variability. For example, of all thalamic nuclei, the ventral medial nucleus 408 had one of the lowest estimated densities of total puncta (19.78 puncta/100 µm2), while the 409 posterior complex had one of the highest estimated densities (110.15 puncta/100 µm2) 410 (Supplementary Table 2). As a validation step, we compared previously published data 411 regarding CA1 (Zhu et al., 2018) with our present data; the total densities of puncta expressing 412 PSD95 and/or SAP102 were remarkably similar (128.95 puncta/100µm2 in previously published 413 data and 130.96 puncta/100µm2 in the present study). 414

415



https://doi.org/10.1101/2020.02.18.953802


16

416

417

418

Figure 7. Densities of PSD95 and SAP102 puncta, and total densities of puncta in different 419 regions of the brain. The total densities of puncta have been calculated from previously 420 published densities and colocalization indexes of PSD95 and SAP102 puncta (Zhu et al., 2018). 421 Symbols represent the different subregions within the major brain regions listed in the legend 422 (see Supplementary Table 2 for the complete list of subregions). The isocortex, the olfactory 423 areas, the hippocampal formation and the cortical subplate have high densities of puncta. The 424 pallidum, the hypothalamus, the brainstem and the cerebellum have low densities of puncta. 425 There is a wide variability in the total densities of puncta in the striatum and thalamus. 426

427

Finally, we estimated the density of synapses expressing PSD95 and/or SAP102, making use of 428 the averaged conversion factor obtained in CA1 (Table 1). The values obtained have been 429 graphically represented in Figure 8. In general, the hippocampal cornu ammonis, the isocortex 430 and the olfactory areas had the highest synaptic densities, intermingled with cortical subplate 431 nuclei. Within the hippocampal formation, the dentate gyrus and the subiculum presented 432 similar densities, but these were lower than in the Ammon’s horn. Striatal nuclei showed 433 considerable variations, but the thalamic nuclei showed the highest variability, as mentioned 434 above. The cerebellar cortex showed homogeneously low densities and the pallidum, 435 hypothalamus and brainstem had the lowest synaptic densities. 436

437

438



https://doi.org/10.1101/2020.02.18.953802


17

439

440

Figure 8. Estimated densities of synapses expressing PSD95 and/or SAP102 in different regions of the 441 brain. The total densities of puncta per 100 square microns have been calculated from previously 442 published densities of PSD95 and SAP102, as well as their colocalization indexes (Zhu et al., 2018). A 443 conversion factor obtained in CA1 was used to estimate the minimum densities of synapses per cubic 444 micron (see text for details). Illustrations and brain regions are based on the Allen Mouse Brain Atlas. 445 The following structures have been labeled: Cerebellum (CBL), CBLCX: cerebellar cortex; DN: Dentate 446 nucleus; IP: Interposed nucleus. Medulla (MD), CO: Cochlear nuclei; PRP: Nucleus prepositus; SPV: 447 Spinal nucleus of the trigeminal; VN: Vestibular nuclei. Midbrain (MB), SN: Substantia nigra. 448 Hypothalamus (HY), ZI: Zona incerta. Thalamus (TH), HA: Habenular nuclei; LG: Lateral geniculate 449 complex; LP: Lateral posterior nucleus; PF: Parafascicular nucleus; PO: Posterior complex; RT: Reticular 450 nucleus; SPF: Suprafascicular nucleus; VM: Ventral medial nucleus; VPL: Ventral posterolateral nucleus; 451 VPM: Ventral posteromedial nucleus. Pallidum, NDB: Diagonal band nucleus; SI: Substantia innominata. 452 Striatum, ACB: Nucleus accumbens; CEA: Central amygdalar nucleus; CP: Caudoputamen; LS: Lateral 453 septal nucleus; MEA: Medial amygdalar nucleus. OT: Olfactory tubercle; SH: septohippocampal nucleus. 454 Cortical subplate, BLA: Basolateral amygdalar nucleus; BMA: Basomedial amygdalar nucleus; CLA: 455 Claustrum; EPd: Endopiriform nucleus, dorsal part; EPv: Endopiriform nucleus, ventral part; LA: Lateral 456 amygdalar nucleus; PA: Posterior amygdalar nucleus. Olfactory areas, AON: Anterior olfactory nucleus; 457 COA: Cortical amygdalar area; PAA: Piriform-amygdalar area; PIR: Piriform area; TR: Postpiriform 458 transition area; TTd: Taenia tecta, dorsal part; TTv: Taenia tecta, ventral part. Hippocampal formation, 459 CA1, CA2, CA3: Cornu Ammonis, fields 1, 2 and 3; DG: Dentate gyrus. ENT: Entorhinal area; SUB: 460 Subiculum. Isocortex, ACAd: Anterior cingulate area, dorsal part; ACAv: Anterior cingulate area, ventral 461 part; AId: Agranular insular area, dorsal part; AIv: Agranular insular area, ventral part; AUD: Auditory 462 areas (d, p, v: dorsal, primary ventral); ECT: Ectorhinal area; GU: Gustatory area; ILA: Infralimbic area; 463 MOp: Primary motor area; MOs: Secondary motor area; ORBl: Orbital area, lateral part; ORBm: Orbital 464 area, medial part; ORBvl: Orbital area, ventrolateral part; PERI: Perirhinal area; PL: Prelimbic area; PTLp: 465 Posterior parietal association areas; RSP: Retrosplenial area; SSp: Primary somatosensory area (bfd, ul 466 m, and n: barrel field, upper limb, mouth and nose representations); TEa: Temporal association areas; 467 VIS: Visual areas (am, p, al: anteromedial, primary, anterolateral). 468

469

470



https://doi.org/10.1101/2020.02.18.953802


18

Discussion 471

In this study, we have —for the first time— analyzed the synaptic density of excitatory and 472 inhibitory synapses, as well as their size, in stratum oriens, stratum radiatum and stratum 473 lacunosum-moleculare of the CA1 hippocampal region of the mouse, using three-dimensional 474 electron microscopy. With this method, long series of consecutive sections are obtained by 475 FIB-SEM, so individual synapses can be unambiguously identified and the number of synapses 476 per unit volume can be directly calculated. However, as with any other electron-microscopy 477 technique, FIB-SEM can only be applied to relatively small regions of tissue, so it is not 478 practical for brain-wide estimations. By contrast, the number of fluorescent puncta expressing 479 PSD95 and/or SAP102 can be quantified brain-wide using SDM, so we have attempted to 480 establish a correlation between the two kinds of measurements. 481

Synaptic sizes and densities in the hippocampus 482

Regarding the size of fluorescent puncta and PSDs, what is actually measured with SDM 483 imaging is different to what is measured after reconstruction of synaptic junctions from serial 484 images obtained by FIB-SEM. In the case of SDM, the images obtained are two-dimensional, so 485 what we actually see is the two-dimensional projection of puncta on the plane of section. 486 Thus, puncta with different orientations will show different apparent surface areas, and only 487 those that are oriented flat with respect to the plane of section will show their true surface 488 area. By contrast, serial images obtained by FIB-SEM allow us to reconstruct the synaptic 489 junctions in 3D. We can then extract the synaptic apposition surface (SAS) from each individual 490 synapse. The SAS represents the surface of apposition between the presynaptic and 491 postsynaptic densities, so the surface area of the SAS is equivalent to the area of the PSD, and 492 we can measure it for every synapse, regardless of its spatial orientation (Morales et al., 2013). 493

We have found that SDM imaging clearly overestimates the size of PSD95 puncta when 494 compared with the actual size of PSDs imaged by FIB-SEM (see Figure 6). This can be due to 495 several factors. Light scatter, glare and blur may contribute to the fact that fluorescent puncta 496 appear to be larger than the actual PSDs. The resolution of SDM is also much lower than the 497 resolution of electron microscopy. In the x-y plane, the resolution of SDM was 84 nm/pixel, 498 while FIB-SEM images were acquired at a resolution of 5 nm/pixel. This makes a pixel area of 499 7056 nm2 for SDM versus only 25 nm2 for FIB-SEM. The lower resolution may result in SDM 500 missing the smaller synapses and those that are oriented perpendicularly to the plane of 501 section. Also, the images of several synapses may overlap throughout the thickness of the SDM 502 optical section. As a result, some puncta may in fact be clusters of two or more synapses. In 503 spite of these differences between the two-dimensional SDM imaging and volume electron 504 microscopy, the measurements of fluorescent puncta by SDM do distinguish the relative size 505 differences between layers or regions, so they are still useful for the identification and 506 classification of synaptic types (Zhu et al., 2018). Both our SDM and FIB-SEM results indicate 507 that excitatory synapses in SLM are larger than in the SR or SO, in line with previous studies in 508 the rat (Megıas et al., 2001). 509

The distribution of synaptic sizes measured from FIB-SEM stacks of images fits a log-normal 510 distribution in the three strata analyzed (see Figure 6 D-F). This trait has also been described in 511 the rat neocortex (Merchán-Pérez et al., 2014; Santuy et al., 2018b). This type of distribution is 512 characterized by a skewed curve with a long tail to the right, and it has been found in other 513 synaptic parameters such as synaptic strength, spike transmission probability, and the size of 514



https://doi.org/10.1101/2020.02.18.953802


19

unitary excitatory postsynaptic potentials (Buzsáki and Mizuseki, 2014; Lefort et al., 2009; Song 515 et al., 2005; Hazan and Ziv, 2020). It is thus tempting to suggest that the size of the synaptic 516 junction is correlated with these and other functional characteristics of the synapse, as has 517 been proposed previously (Santuy et al., 2018b). 518

Regarding the densities of puncta and synapses, previous studies in the rat SR reported 2.2 519 synapses/µm3 using EM and three-dimensional reconstructions (Mishchenko et al., 2010), and 520 similar estimates using stereological methods (Sorra et al., 1998). In both cases, the reported 521 synapse densities were lower than the density we have found in the mouse SR (2.4 522 synapses/µm3). Differences between species may explain the discrepancies, although we 523 cannot rule out the possibility of other sources of bias, such as the different methods used. On 524 the other hand, our SDM results regarding the total density of PSD95 and SAP102 puncta in the 525 hippocampus were very similar to previously reported data (Zhu et al., 2018). We also provide 526 information about the amount of inhibitory synapses, represented by symmetric or type 2 527 synapses (Colonnier, 1968; Gray, 1959). These do not express PSD95 or SAP102, so their 528 densities cannot be estimated from our SDM data. However they can be identified in FIB-SEM 529 images because of their thin PSD (Merchán-Pérez et al., 2009). In our CA1 samples, symmetric 530 synapses represented 4.4% of the total number of synapses. This is in line with results in SR 531 and SO of the rat CA1, where percentages of inhibitory synapses as low as 3% have been 532 reported in thin dendrites, which predominate in our samples (Megıas et al., 2001). 533 Interestingly, they also reported that —in line with our results— SLM had the highest 534 percentage of inhibitory synapses (leaving aside the stratum pyramidale and the thick proximal 535 dendrites, which were not included in our study). 536

Brain-wide estimations of the minimum densities of synapses 537

We next applied a conversion factor obtained in the hippocampus to calculate synaptic 538 densities brain-wide. The conversion factor was calculated as the ratio between the densities 539 of excitatory (asymmetric) synapses obtained by FIB-SEM and the total density of PSD95 and 540 SAP102 puncta obtained by SDM. We found that the conversion factors were very similar in 541 the three CA1 layers studied, and we used an averaged conversion factor for brain-wide 542 estimations (see Table 1). It is important to bear in mind the limitations of this procedure to 543 ensure that the results are interpreted correctly. 544

While it is clear that only excitatory glutamatergic synapses express PSD95 and/or SAP102 545 (Nithianantharajah et al., 2013; Frank et al., 2017; Zhu et al., 2018), the question of whether all 546 excitatory synapses express these scaffolding proteins does not have a simple answer. In the 547 adult mouse hippocampus, it has been recently claimed that all Schaffer 548 collateral/commissural synapses in the SR of CA1 show immunogold staining for PSD95 549 (Yamasaki et al., 2016). This is probably an overestimate, since our own data indicate that 550 there is a population of synapses that do not express PSD95, but do express SAP102 551 (Supplementary Table 2). In any case, if we consider that Schaffer collateral/commissural fibers 552 are the origin of the vast majority of synapses in SR and SO, we can assume that most, if not 553 all, synapses in these strata express PSD95, SAP102 or a combination of the two. It is likely to 554 be the same case in SLM, since the ratio between the number of fluorescent puncta and the 555 actual density of synapses measured by FIB-SEM is very similar to that of the two other layers 556 (see Table 1). Lower percentages of immunolabeling of synapses with PSD95 have been 557 reported in the rat hippocampus (Sans et al., 2000), but this has been attributed to the low 558 sensitivity of the technique (Sassoé-Pognetto et al., 2003). In our case, the advantage of the 559



https://doi.org/10.1101/2020.02.18.953802


20

genetic labeling method is that all PSD95 and SAP102 proteins are labeled, so a more reliable 560 detection is to be expected. 561

However, even if we assume that the vast majority of excitatory synapses in CA1 express 562 PSD95 and/or SAP102, and that we can detect them in a reliable way, the question remains as 563 to whether this would be the case in other brain regions. For example, PSD95 was regarded as 564 “a fundamental structural component of most, if not all, excitatory PSDs isolated from the rat 565 cerebral cortex” (Petersen et al., 2003). Other studies seem to confirm this view (Swulius et al., 566 2010; DeGiorgis et al., 2006), while lower percentages of PSD95-expressing synapses have also 567 been reported (Aoki et al., 2001; Farley et al., 2015). Brain-wide studies in the mouse seem to 568 confirm that the abundance of scaffolding proteins like PSD95 and SAP102 differs depending 569 on the brain area (Roy et al., 2018; Zhu et al., 2018). 570

Therefore, to interpret our results correctly, we must clearly assume that not all excitatory 571 synapses throughout the brain express PSD95 and/or SAP102. Since our calculations are based 572 only on the population of synapses that express these scaffolding proteins, our estimations of 573 synaptic densities do in fact underestimate the actual densities of excitatory synapses. In other 574 words, our estimations represent the lower boundary of the densities of excitatory synapses in 575 different brain regions. The upper boundary cannot be estimated from our present data, since 576 this would require knowing the proportion of excitatory synapses that do not express PSD95 577 and/or SAP102 in each brain region. 578

Only a systematic exploration of the different regions of the brain with FIB-SEM or similar 579 methods will settle the possible discrepancies between our present estimations and the actual 580 values. However, we can compare our estimations with previous studies, when available. In 581 the mouse neocortex, previously reported synaptic densities using different stereological 582 methods were either lower (Sadaka et al., 2003; Schüz and Palm, 1989) or higher (DeFelipe et 583 al., 1997) than our present estimations. In the juvenile rat somatosensory cortex, the mean 584 density of synapses in the neuropil has been reported to be between 0.87 and 0.89 585 synapses/µm3 using FIB-SEM (Anton-Sanchez et al., 2014; Santuy et al., 2018a), which is below 586 our present estimation for the adult mouse somatosensory cortex (1.4 to 1.9 synapses/µm3, 587 see Supplementary Table 2). However, these differences may be due to species and/or age 588 differences (e.g., DeFelipe et al., 1997). In the rat cerebellum, the density of synapses has been 589 previously reported to be 0.8 synapses/µm3 in the molecular layer (Napper and Harvey, 1988), 590 while our present estimations for the mouse cerebellar cortex range from 0.5 to 0.6 591 synapses/µm3 (Supplementary Table 2). Therefore, we currently lack data that are directly 592 comparable to our present estimations, since methodological bias is probably at play in those 593 cases, leaving aside the possible species and age differences. Although work is already in 594 progress on the mouse somatosensory cortex using a FIB-SEM methodology that is similar to 595 the one presented here, it would not be practical to wait until results from even a fraction of 596 the 113 subregions examined here become available. Therefore, our calculations must be 597 regarded as reasonable —but provisional— estimations of the minimum densities of 598 glutamatergic synapses in the different brain regions. 599

In summary, it is important to emphasize that acquiring multiple samples at different scales is 600 a highly effective way to obtain a dataset that allows comprehensive analysis of the brain. 601 Since the whole brain cannot be fully reconstructed at the ultrastructural level, it seems clear 602 that only by combining studies at the meso- and nano-scopic levels (light and electron 603 microscopy) can we fully understand the structural arrangement of the brain as a whole [see, 604 for example (Markram et al., 2015, Kashiwagi et al., 2019)]. Using this strategy, we provide an 605



https://doi.org/10.1101/2020.02.18.953802


21

estimation of the minimum densities of glutamatergic synapses in the different brain regions. 606 These data, in combination with previous studies on the relationship between the connectome 607 and synaptome (Zhu et al., 2018), can be used to identify common and differing principles of 608 synaptic organization. This in turn could serve to further advance efforts to validate and refine 609 realistic brain models. 610

611

612

613

Acknowledgements 614

This work was supported by grants from the following entities: the Spanish “Ministerio de 615 Ciencia, Innovación y Universidades” (grant PGC2018-094307-B-I00 and the Cajal Blue Brain 616 Project [C080020-09; the Spanish partner of the Blue Brain Project initiative from EPFL, 617 Switzerland]; the European Union’s Horizon 2020 Research and Innovation Programme under 618 grant agreement No. 785907 (Human Brain Project, SGA2); the Wellcome Trust (Technology 619 Development Grant 202932); and the European Research Council (ERC) under the European 620 Union’s Horizon 2020 research and innovation programme (695568 SYNNOVATE). L.T.-R. is a 621 recipient of grants from the EMBO Long-term fellowship 2016–2018 and the IBRO-PERC 622 InEurope grants programme. 623

624



https://doi.org/10.1101/2020.02.18.953802


22

References 625

Anton-Sanchez L, Bielza C, Merchan-Perez A, Rodríguez J-R, DeFelipe J, Larrañaga P (2014) Three-626 dimensional distribution of cortical synapses: a replicated point pattern-based analysis. Front 627 Neuroanat 8:85. 628

Aoki C, Miko I, Oviedo H, Mikeladze-Dvali T, Alexandre L, Sweeney N, Bredt DS (2001) Electron 629 microscopic immunocytochemical detection of PSD-95, PSD-93, SAP-102, and SAP-97 at 630 postsynaptic, presynaptic, and nonsynaptic sites of adult and neonatal rat visual cortex. 631 Synapse 40:239–257. 632

Buzsáki G, Mizuseki K (2014) The log-dynamic brain: how skewed distributions affect network 633 operations. Nat Rev Neurosci 15:264–278. 634

Chen X, Winters C, Azzam R, Li X, Galbraith JA, Leapman RD, Reese TS (2008) Organization of the core 635 structure of the postsynaptic density. PNAS 105:4453–4458. 636

Chen X, Winters C, Crocker V, Lazarou M, Sousa AA, Leapman RD, Reese TS (2018) Identification of PSD-637 95 in the Postsynaptic Density Using MiniSOG and EM Tomography. Front Neuroanat 12. 638

Colonnier M (1968) Synaptic patterns on different cell types in the different laminae of the cat visual 639 cortex. An electron microscope study. Brain Res 9:268–287. 640

DeFelipe J, Fairén A (1993) A simple and reliable method for correlative light and electron microscopic 641 studies. J Histochem Cytochem 41:769–772. 642

DeFelipe J, Marco P, Busturia I, Merchán-Pérez A (1999) Estimation of the number of synapses in the 643 cerebral cortex: methodological considerations. Cereb Cortex 9:722–732. 644

DeFelipe J, Marco P, Fairén A, Jones EG (1997) Inhibitory synaptogenesis in mouse somatosensory 645 cortex. Cereb Cortex 7:619–634. 646

DeGiorgis JA, Galbraith JA, Dosemeci A, Chen X, Reese TS (2006) Distribution of the scaffolding proteins 647 PSD-95, PSD-93, and SAP97 in isolated PSDs. Brain Cell Biology 35:239–250. 648

Farley M, Swulius M, Waxham M (2015) Electron Tomographic Structure and Protein Composition of 649 Isolated Rat Cerebellar, Hippocampal and Cortical Postsynaptic Densities. Neuroscience 650 304:286–301. 651

Frank RAW, Zhu F, Komiyama NH, Grant SGN (2017) Hierarchical organization and genetically separable 652 subfamilies of PSD95 postsynaptic supercomplexes. J Neurochem 142:504–511. 653

Gray EG (1959) Axo-somatic and axo-dendritic synapses of the cerebral cortex: an electron microscope 654 study. J Anat 93:420–433. 655

Hafner A-S, Penn AC, Grillo-Bosch D, Retailleau N, Poujol C, Philippat A, Coussen F, Sainlos M, Opazo P, 656 Choquet D (2015) Lengthening of the Stargazin Cytoplasmic Tail Increases Synaptic 657 Transmission by Promoting Interaction to Deeper Domains of PSD-95. Neuron 86:475–489. 658

Hazan L, Ziv NE (2020) Activity dependent and independent determinants of synaptic size diversity. J 659 Neurosci 40:2828–2848 660

Howard CV, Reed MG (2005) Unbiased stereology : three-dimensional measurement in microscopy, 2nd 661 ed. ed. Oxon, UK: Garland Science/BIOS Scientific Publishers. 662

Husi H, Ward MA, Choudhary JS, Blackstock WP, Grant SGN (2000) Proteomic analysis of NMDA 663 receptor–adhesion protein signaling complexes. Nat Neurosci 3:661–669. 664

Jeyifous O, Lin EI, Chen X, Antinone SE, Mastro R, Drisdel R, Reese TS, Green WN (2016) Palmitoylation 665 regulates glutamate receptor distributions in postsynaptic densities through control of PSD95 666 conformation and orientation. PNAS 113:E8482–E8491. 667

Kashiwagi Y, Higashi T, Obashi K, Sato Y, Komiyama NH, Grant SGN, Okabe S (2019) Computational 668 geometry analysis of dendritic spines by structured illumination microscopy. Nat Commun 669 10:1285. 670

Lefort S, Tomm C, Floyd Sarria J-C, Petersen CCH (2009) The Excitatory Neuronal Network of the C2 671 Barrel Column in Mouse Primary Somatosensory Cortex. Neuron 61:301–316. 672

Levy JM, Chen X, Reese TS, Nicoll RA (2015) Synaptic Consolidation Normalizes AMPAR Quantal Size 673 following MAGUK Loss. Neuron 87:534–548. 674

Markram H et al. (2015) Reconstruction and Simulation of Neocortical Microcircuitry. Cell 163:456–492. 675 Megıas M, Emri Z, Freund TF, Gulyás AI (2001) Total number and distribution of inhibitory and excitatory 676

synapses on hippocampal CA1 pyramidal cells. Neuroscience 102:527–540. 677



https://doi.org/10.1101/2020.02.18.953802


23

Merchán-Pérez A, Rodriguez J-R, Alonso-Nanclares L, Schertel A, Defelipe J (2009) Counting synapses 678 using FIB/SEM microscopy: a true revolution for ultrastructural volume reconstruction. Front 679 Neuroanat 3:18. 680

Merchán-Pérez A, Rodríguez J-R, González S, Robles V, Defelipe J, Larrañaga P, Bielza C (2014) Three-681 dimensional spatial distribution of synapses in the neocortex: a dual-beam electron microscopy 682 study. Cereb Cortex 24:1579–1588. 683

Migaud M, Charlesworth P, Dempster M, Webster LC, Watabe AM, Makhinson M, He Y, Ramsay MF, 684 Morris RGM, Morrison JH, O’Dell TJ, Grant SGN (1998) Enhanced long-term potentiation and 685 impaired learning in mice with mutant postsynaptic density-95 protein. Nature 396:433–439. 686

Mishchenko Y, Hu T, Spacek J, Mendenhall J, Harris KM, Chklovskii DB (2010) Ultrastructural analysis of 687 hippocampal neuropil from the connectomics perspective. Neuron 67:1009–1020. 688

Morales J, Alonso-Nanclares L, Rodríguez J-R, DeFelipe J, Rodríguez Á, Merchán-Pérez A (2011) Espina: a 689 tool for the automated segmentation and counting of synapses in large stacks of electron 690 microscopy images. Front Neuroanat 5:18. 691

Morales J, Rodríguez A, Rodríguez J-R, DeFelipe J, Merchán-Pérez A (2013) Characterization and 692 extraction of the synaptic apposition surface for synaptic geometry analysis. Front Neuroanat 693 7:20. 694

Napper RMA, Harvey RJ (1988) Number of parallel fiber synapses on an individual Purkinje cell in the 695 cerebellum of the rat. J Comp Neurol 274:168–177. 696

Nithianantharajah J, Komiyama NH, McKechanie A, Johnstone M, Blackwood DH, St Clair D, Emes RD, 697 van de Lagemaat LN, Saksida LM, Bussey TJ, Grant SG (2013) Synaptic scaffold evolution 698 generated components of vertebrate cognitive complexity. Nat Neurosci 16:16–24. 699

Paxinos G, Franklin KB (2004) The mouse brain in stereotaxic coordinates. Gulf Professional Publishing. 700 Petersen JD, Chen X, Vinade L, Dosemeci A, Lisman JE, Reese TS (2003) Distribution of Postsynaptic 701

Density (PSD)-95 and Ca2+/Calmodulin-Dependent Protein Kinase II at the PSD. J Neurosci 702 23:11270–11278. 703

Qiu Z, Yang L, Lu W (2012) A new feature-preserving nonlinear anisotropic diffusion for denoising images 704 containing blobs and ridges. Pattern Recognition Letters 33:319–330. 705

Roy M, Sorokina O, McLean C, Tapia-González S, DeFelipe J, Armstrong J, Grant S (2018) Regional 706 Diversity in the Postsynaptic Proteome of the Mouse Brain. Proteomes 6:31. 707

Sadaka Y, Weinfeld E, Lev DL, White EL (2003) Changes in mouse barrel synapses consequent to sensory 708 deprivation from birth. J Comp Neurol 457:75–86. 709

Sans N, Petralia RS, Wang Y-X, Blahos J, Hell JW, Wenthold RJ (2000) A Developmental Change in NMDA 710 Receptor-Associated Proteins at Hippocampal Synapses. J Neurosci 20:1260–1271. 711

Santuy A, Rodriguez J-R, DeFelipe J, Merchan-Perez A (2018a) Volume electron microscopy of the 712 distribution of synapses in the neuropil of the juvenile rat somatosensory cortex. Brain Struct 713 Funct 223:77–90. 714

Santuy A, Rodriguez J-R, DeFelipe J, Merchan-Perez A (2018b) Study of the Size and Shape of Synapses in 715 the Juvenile Rat Somatosensory Cortex with 3D Electron Microscopy. eNeuro 5:e0377-17.2017. 716

Sassoé-Pognetto M, Utvik JK, Camoletto P, Watanabe M, Stephenson FA, Bredt DS, Ottersen OP (2003) 717 Organization of postsynaptic density proteins and glutamate receptors in axodendritic and 718 dendrodendritic synapses of the rat olfactory bulb: PSD-95/PSD-93 in the Olfactory Bulb. J 719 Comp Neurol 463:237–248. 720

Schüz A, Palm G (1989) Density of neurons and synapses in the cerebral cortex of the mouse. J Comp 721 Neurol 286:442–455. 722

Song S, Sjöström PJ, Reigl M, Nelson S, Chklovskii DB (2005) Highly Nonrandom Features of Synaptic 723 Connectivity in Local Cortical Circuits. PLoS Biology 3:e68. 724

Sorra KE, Fiala JC, Harris KM (1998) Critical assessment of the involvement of perforations, spinules, and 725 spine branching in hippocampal synapse formation. J Comp Neurol 398:225–240. 726

Swulius MT, Kubota Y, Forest A, Waxham MN (2010) Structure and composition of the postsynaptic 727 density during development. J Comp Neurol 518:4243–4260. 728

Valtschanoff JG, Burette A, Wenthold RJ, Weinberg RJ (1999) Expression of NR2 receptor subunit in rat 729 somatic sensory cortex: Synaptic distribution and colocalization with NR1 and PSD-95. Journal 730 of Comparative Neurology 410:599–611. 731

Yamasaki M, Fukaya M, Yamazaki M, Azechi H, Natsume R, Abe M, Sakimura K, Watanabe M (2016) 732 TARP γ-2 and γ-8 Differentially Control AMPAR Density Across Schaffer Collateral/Commissural 733 Synapses in the Hippocampal CA1 Area. J Neurosci 36:4296–4312. 734



https://doi.org/10.1101/2020.02.18.953802


24

Ye F, Zeng M, Zhang M (2018) Mechanisms of MAGUK-mediated cellular junctional complex 735 organization. Curr Opin Struct Biol 48:6–15. 736

Zhu F, Cizeron M, Qiu Z, Benavides-Piccione R, Kopanitsa MV, Skene NG, Koniaris B, DeFelipe J, Fransén 737 E, Komiyama NH, Grant SGN (2018) Architecture of the Mouse Brain Synaptome. Neuron 738 99:781-799.e10. 739

Zhu J, Shang Y, Zhang M (2016) Mechanistic basis of MAGUK-organized complexes in synaptic 740 development and signalling. Nat Rev Neurosci 17:209–223. 741

742



https://doi.org/10.1101/2020.02.18.953802


Documents

Estimation of the number of synapses in the hippocampus ...Feb 18, 2020 · 1 1 Estimation of the number of synapses in the hippocampus and 2 brain-wide by volume electron microscopy